504 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



angle is the apex of the interbrachial angle and the arms are 



the first marginal plates (figs. 3 and 3a, pi. 156) ; proximal 



adambulacral plates not higher than long ; gonads 2 or 4 to each 



ray. Genotype, Brisinga panamensis LiuAwig-Astrolirus Fisher.* 



e^ Abactinal skeleton of genital region of ray composed of a uniform 



armor of thin spiniferous, more or less overlapping plates, but 



not of independent arches or costae. 



f. Primary plates of disk much larger than the others which are 



small ; gonads 2 to each ray ; an interradial plate, vertical in 



position, is interpolated between the proximal ends of the first 



adambulacral plates, and touches the mouth plates, but does not 



encroach upon the actinal surface as In Colpaster. Genotype, 



Belgicclla racovltzana Ludwig Belgicella. Ludwig. 



f. Primary plates of disk not conspicuous, and not distinguishable 



from the other abactinal plates of disk. 



g^. An azygous interradial plate, shield-shaped and conspicuous 



superficially on the actinal surface, separates the first pair 



of free adambulacral plates; gonads unknown. Genotype, 



Colpaster scutigerulus Sladen Colpaster Sladen. 



g\ First pair of adambulacral plates not separated by an azygous 

 plate, but united as in Brisinga, and above them is a pair 

 of conspicuous, united, first marginal plates — 4 in all in each 

 interradial angle ; gonads numerous, in a series along either 

 side of the ray. Genotype, Freyclla fecunda Fisher. 



FreyeUastcr Fisher. 

 d'. No syzygy, or nonmuscular symphysis, between the first and second 

 adambulacral plates nor between the upper part of the second and 

 third ambulacral plates, but a normal muscular symphysis not 

 different from the others ; no marginal plates directly above the first 

 pair of adambulacral plates ; the latter are usually not united except 

 sometimes by the proximal ends ; touching the lower end of the inter- 

 radial plate is a pair of very inconspicuous plates lying in the same 

 plane, and superficially appearing to be a part of the interradial 

 plate.^ These are really the outer ends of the mouth plates which 

 project dorsally behind the first pair of adambulacral plates. The 

 latter, by sometimes touching at their inner ends or apposing their 

 lateral faces, segregate this dorsal portion of the mouth plates from 

 the actinal, spine-bearing part. For this reason they may easily be 

 mistaken for 2 entirely independent plates (see fig. 4, pi. 1.56). 

 e\ Abactinal skeleton of ray composed of a uniform armor of thin, 

 spiniferous more or less overlapping plates as in Freyellaster, not 

 of spaced, independent arches or costae; furrow spinelets not 

 modified or expanded at the tip. Genotype, Freyella spinosa 



Perrier Freyclla Perrier, 



e'. Abactinal skeleton of ray composed of independent transverse 

 arches, composed, on radial area, of flattened, overlapping plates 

 (not of elongate, narrow, more or less compressed ossicles as in 

 Brisinga) ; these costae bear spinelets and are separated by inter- 

 vals devoid of plates ; furrow spinelets with curiously modified 

 expanded tips. Genotype, Astrocles actinodetus—Astroclcs Fisher.' 



1 Fisher. 1917f, pp. 424, 428. 



- The interradial plate described by Ludwig in Belgicella is very probably the true 

 Interradial plate plus these extreme outer ends of the mouth plates, which, unless treated 

 U'lth potash, appear to be a part of the Interradial plate (Ludwig, 190.3, p. 60). 



3 Fisher, 1917f, pp. 426. 430. 



